Accuracy of Forecasts

Can anybody please vouch for the accuracy of the forecasts on this J2SKI website? For example today (10/4/07) I am looking at the forecast for the next 7 days in Val d'Isere and it shows a lot of snow falling over these days, almost every day including the whole of this current week. Whereas having just looked at several other ski forecast sites they are prediciting just light snow and rain - hard to believe they are talking about the same resorts! So does anyone have any info as to how accurate the forecasts on this site are please?

They are in my experience of Austria and Scotland this year and last, wildly inaccurate and extremely optimistic.
I am not sure how they do it but both this site and snow-forecast.com seem to have a fantasy factor built in to their forecasts which increases the further into the future the forecast is for, this gradually reduces as it becomes closer to the day.
I am sure if you check the following days forecast every day for the next 6 days it will probably be accurate -ish.
But it isn't just the forecasts that exaggerate the snow depths, the resorts lie in their snow reports too.
Anyone reading snow reports and weather forecasts now would think we had entered the next ice age not and not worrying about global warming.

Its the mountains! Anything beyond 2/3 days is, to put it mildly, highly speculative, particularly when it is changeable. I think you will generally find that most forecasts are fairly close and reasonably reliable in their 2/3 day forecasts.

24 hours max for a forecast, anything more is fantasy.

Just to pick up on this... we certainly aim for accuracy, but we are at the mercy both of the extreme variability inherent in mountain weather and the weaknesses in the forecast models we use.

twoshortplanks put his/her finger on one of the apparent GFS (Global Forecast System) model deficiencies - deviations in the model seem to be sometimes compounded with time-to-go; so the estimates of precipitation sometimes start out large (at 7 days) and fall progressively as the actual day approaches.

This effect seems to apply mostly to small snowfalls - the combination of our processing and the GFS model generally predicts the big dumps quite well, even from a week out.

We will be looking at this over the summer and looking at ways to minimise such deviations with our own post-processing of the raw GFS data.

Purely anecdotally, this last week we were out in Les Arcs; several minor snowfalls occurred throughout the week and our forecast generally over-stated these (e.g. predicting 5cm when 1cm actually fell). The major dump of the week (30cm+) on Friday was in our forecast from 7 days out, however, although the estimate bounced around between 20cm and 35cm as the day approached.

We've archived a huge amount of data from this season and will be crunching through it over the summer to compare the forecasts with actual reported snowfall... sad but true.

So when you're sunning yourself on a beach this summer, spare a thought for us snow-nerds slaving over hot servers and rarely seeing daylight... kinda like bears hibernating but with the seasons switched... 8)

Admin wrote:So when you're sunning yourself on a beach this summer, spare a thought for us snow-nerds slaving over hot servers and rarely seeing daylight... kinda like bears hibernating but with the seasons switched... 8)

God bless you 8)

For skiing purposes, it doesn't matter if there is going to be 2cm or 8 cm, from day to day. It matters if there is 30cm or more, particularly when sustained over a few days.

Much more important is the sustaining of low temperatures, obviously for the retention and quality of new snow, but also, in retaining the base, which in most winters, below 2000mtrs, is provided by artificial snow.

Over the years, artificial snow production and piste machine management has improved. This has the benefit of providing good base for the new snow, and for extending the season.

Yes, I look along the line of snow drops, but I also focus along the mountain temperature line.

Dave Mac wrote:
Much more important is the sustaining of low temperatures, obviously for the retention and quality of new snow, but also, in retaining the base, which in most winters, below 2000mtrs, is provided by artificial snow.

Over the years, artificial snow production and piste machine management has improved. This has the benefit of providing good base for the new snow, and for extending the season.

Really good points Dave. A good artificial base can stand a lot of weather abuse. A few -20c nights and some good grooming can bring it back from an icey mess. I've really started watching humidity. The base can take warm weather so long as it's not too humid.

Trencher